A transformer is the most important equipment for power supply to consumers. Failure of one power transformer may cause long interruptions in supply, costly repairs and loss of revenue. For uninterrupted power supply to consumers, proper maintenance, particularly preventive maintenance, is very necessary. The failure in magnetic, electric and dielectric circuits as well as structural failure may cause extensive damage to the equipment and surroundings. Proper operation and maintenance procedure may help to prevent failure and extend life of operation of the transformer.

The mineral oil uses in transformers, acts as a dielectric medium and also as a heat transfer agent. The breakdown of electrical insulating materials and related components inside the transformer liberates gases within the unit. The distribution of these gases can be related to the type of electrical fault, and the rate of gas generation can indicate the severity of the fault. The identity of the gases being generated by a particular unit can be very useful information in any preventive maintenance program [1] - [3] . There are several techniques in detecting those fault gases and DGA was recognized as the most informative method. This method involves sampling of the oil to measure the concentration of the dissolved gases. The risk of failure of transformers in service, based on dissolved gas analysis (DGA), depends on three main parameters: the type of fault involved, the location of the fault (in oil or in paper), and the amount of gases formed (concentrations and rates).

・ The most dangerous faults are: high-energy arcing faults in oil and paper (D2), low-energy arcing faults in paper (D1), and hot spots in paper of high temperatures (>700˚C T3 and 300˚C - 700˚C T2).

・ Less dangerous faults are: low-energy arcing faults in oil (D1), hot-spots in oil (>700˚C T3 and 300˚C - 700˚C T2), and hot spots in paper of low temperature (<300˚C T1).

・ Non-dangerous faults are: hot spots in oil (<300˚C T1), producing only “stray gassing” of oil, corona partial discharges (PD) (unless very high levels of hydrogen are formed), catalytic reactions with water, and aging of paper.

A risk model for transformer failures [4] - [6] , based on aging, can be represented by Perks formula:

Figure 1 is the corresponding exponential curve for 50% failure rate at the age of 50 year. To include the random events (lighting, vandalism, etc.) separate from the aging component, the constant “A” is set at 0.005. In Figure 2 is represented the age of the main transformers of Albanian Transmission System. We can see that for some transformers the failure rate is higher than 10%. They are expensive equipments; that’s why there has been a growing interest in the technique to diagnose, determine and decide the condition assessment of transformer insulation [7] [8] .

The general type of fault (PD, D1, D2, T1, T2, and T3) can be identified by several methods: e.g., Rogers, Key Gas, IEC ratios, and Duval Triangle [9] [10] . Electrical and thermal stresses such as arching, partial discharges and overheating cause degradation of dielectric oil and solid dielectric cellulose materials.

The degradation of insulation produces different gases. Different degradation mechanisms generate different gases thus making it possible to determine the degrading part of the transformer. Power transformer gas-in-oil analysis (DGA) can be used for effective diagnostics and monitoring. The IEEE and IEC guides [11] [12] offer a variety of ratio-based tools to diagnose DGA data. A quick summary of the different tools found in the current IEEE guide and the IEC guide is in Table 1.

DGA diagnostic tools vary in their complexity and accuracy. Table 2 shows the comparison among the most widely used DGA diagnostic methods.

The Albanian Power Companies started to invest to DGA laboratory tests from two years, so there is no previous dissolved gas history data. In IEC Standards 60599 is suggested a four-level criterion to classify risks to transformers evaluating individual gas and TDCG concentrations when there is no previous dissolved gas history. That is why among different methods for interpreting DGA results, we will use hybrid method based in the Key Gas Method with total dissolved combustible gas method to be evaluated and then verified by the Rogers’ Ratios.

Typical gases generated from mineral oil/cellulose (paper and pressboard) insulated transformers include: Hydrogen, H₂; Methane, CH4; Ethane, C₂H₆; Ethylene, C₂H₄; Acetylene, C₂H₂; Carbon Monoxide, CO; Carbon Dioxide, CO₂. Additionally, oxygen and nitrogen are always present; their concentrations vary with the type of preservation system used on the transformer. In addition, gases such as propane, butane and others can be formed as well, but their use for diagnostic purposes is not widespread. The concentration of the different gases provides information about the type of incipient-fault condition present as well as the severity. In Table 3 and Table 4 are shown fault interpretations from dissolved gases according [13] .

The key gas method is mainly depends on the quantity of fault gases release in mineral oil when fault occur.

Fault gases are caused by corona (partial discharge), thermal heating and arcing. The Key Gas Method considers the following four general fault types:

1) Thermal fault due to overheated oil;

2) Thermal fault due to overheated cellulose;

3) Electrical fault due to corona;

4) Electrical fault due to arcing.

The standard of IEEE Std C57.104-1991 indicates the key gases and their relative proportions for four fault types (Table 5). Generally, the thermal decomposition of oil produced more than 60% of ethylene (C₂H₄) and thermal decomposition of cellulose produce key gas carbon monoxide (CO) is 90%. In case of corona in oil mainly produce large amount principal gas hydrogen nearly 80% and due to arcing key gas acetylene produced 30% with trace quantity of hydrogen.

The total dissolved combustible gas (TDCG) is the definition of the sum of the combustible gas concentrations as follows:

InIEC Standards 60599 it is suggested a four-level criterion (Table 6) to classify risks to transformers when there is no previous dissolved gas history as follow:

Condition 1: TDGC below this level indicates the transformer is operating satisfactorily;

Condition 2: TDGC within this range indicates greater than normal combustible level. Action should be taken to establish a trend;

Condition 3: TDGC within this range indicates a high level of decomposition. Immediate action should be taken to establish a trend;

Condition 4: TDGC within this range indicates excessive decomposition. Continued operation could result in failure of the transformer.

This method uses five gases (H₂, C₂H₆, CH4, C₂H₂ and C₂H₄) to generate codes based on their composition ratio [14] . The codes are then used to categorize a range of ratios used to diagnose the fault as shown in Table 7.

The method utilizes the ratio of gas concentration to indicate fault types. Rogers Ratio method can be used when any of individual gases exceeds its normal limit and it does not depend on specific gas concentrations. The Rogers Ratio method utilizes four ratios; C₂H₆/CH₄, C₂H₂/C₂H₄, CH₄/H₂, and C₂H₄/C₂H₆ that leads to twenty-two proposed diagnosis as shown in Table 8 [15] . Rogers Ratio method diagnosis provides more interpretation details in terms of temperature range of decomposition.

In [16] , the Rogers Ratio method diagnosis was revised, as shown in Table 9, to include only six diagnosis interpretations; normal, low-energy density arcing-PD, arcing-high-energy discharge (D2), low temperature thermal (T1), thermal 300˚C - 700˚C (T2) and thermal >700˚C (T3).

This method does not consider dissolved gases below normal concentration limits; also certain ratio values are inconsistent with the diagnostic assigned and lead to invalid codes [17] .

The hybrid method proposed in this paper is a step-by-step procedure (flow chart Figure 3) to diagnose faults

using the Key Gas Method with total combustible gas method (TCGM) and then verified by the Rogers’ Ratios.

Step 1. Gas concentrations are obtained by DGA and TCGM is evaluated;

Step 2. If at least one of the gas concentrations (in ppm) for H₂, CH₄, C₂H₂, C₂H₄, C₂H₆ and TCGM exceeds the values for abnormal limit (see Table 3), the unit is considered faulty; proceed to Step 4;

Step 3. If at least one of the gas concentrations (in ppm) for H₂, CH₄, C₂H₂, C₂H₄, C₂H₆ and TCGM exceeds the values for limit C₂ (see Table 6), the unit is considered faulty; precede to Step 4, otherwise “No fault condition”;

Step 4. Determining the values of ratio obtained from Table 7 in the order of ratio CH₄/H₂, C2H₂/CH₄, C₂H₄/C₂H₆ and C₂H6/CH₄ procedure. Define the codes from Table 7;

Step 5. Each successive code is compared to the values obtained from Table 8 to define the fault;

Step 6. If all succeeding ratios for a specific fault type fall within the values (column) given in Table 9, the type of fault is defined;

Step 7. The gas concentrations for H₂, CH₄, C₂H₂, C₂H₄, C₂H₆ in % of TCGM is evaluated; If relative proportions of the keys gases for a specific fault type fall within the values given in Table 5, the suggested diagnosis is valid. The report of fault is defined.

The method was tested against all the 30 cases in the data set [18] . In Table 10, we show a set of data that illustrate the method as well as the evaluation of the key gases concentrations according Table 6 (Step 1).

In Table 11 are represented the result of application of step 2 and 3 of above methods and the first step evaluation of faults. In Table 12 are represented the Rogers’ Ratios and the fault according to the Rogers Code Method.

Table 12 represented the result of application of step 4 and 5 of above methods based in the following reasoning:

1) Since total combustible gases (TDCG) are LESS the 720 ppm, the Rogers’ Ratios result indicated below is considered less significant. Since C₂H₄ is ABNORMAL a Core & Tank circulating currents overheated joints;

2) Since total combustible gases (TDCG) are GREATER than 720 ppm, the Rogers’ Ratios result indicated below should be considered more significant. Winding circulating currents;

3) Since total combustible gases (TDCG) are GREATER than 720 ppm, the Rogers’ Ratios result indicated below should be considered more significant. Arc, with power follow through;

4) Since total combustible gases (TDCG) are GREATER than 720 ppm, the Rogers’ Ratios result indicated below should be considered more significant. Core & Tank circulating currents overheated joints;

5) Since total combustible gases (TDCG) are GREATER than 720 ppm, the Rogers’ Ratios result indicated below should be considered more significant. Arc, with power follow through.

From Tables 10-12, we can see the evaluation of the results for 5 cases has given a successful prediction with the Key Gas Method with total combustible gas method (TCGM) for cases 2 - 5 and a successful prediction with the Rogers’ Ratios for case 2.

The evaluation of the results for 30 cases in the data set give a successful prediction of the Key Gas Method with total combustible gas method (TCGM) of about 78%, a successful prediction of the Rogers’ Ratios of about 45% and the successful prediction of the Hybrid Method of about 87%.

The Albanian Power Companies started to invest in the last years to DGA laboratory tests to monitor main HPPs transformer units [19] . The data consists of numerical measurement of each individual key gas in ppm. In Table 13 are represented the data measured in transformer of HEC Koman, one of the main power plant in Albania in two consecutive months.

By looking at the relative proportions of gases in the DGA results and the monthly increase it could be possible to identify the type of fault occurring in the transformer. It is evident an increase of Carbon Dioxide, Oxygen and humidity, while TDCG are LESS the 720 ppm. So, the rehabilitation of oil through degasification and dehumidification was necessary. After this process, as presented in Table 14, the relative proportions of gases in the DGA results at normal value.

For uninterrupted power supply to consumers, proper maintenance of transformer, particularly preventive maintenance, is very necessary. The IEC standards 60599 and IEEE standards C57.104-1991 help in interpretation of DGA results, which together with further field inspections and experienced personnel judgment can give an answer to what is going on inside a suspected unhealthy transformer. The Key Gas Method in principal gives no answer in multiple fault condition.

To increase the accuracy of the interpretation and the certainty of the transformer condition, a hybrid method based on the Key Gas Method with total combustible gas method (TCGM) verified by the Rogers’ Ratios is used to improve the accuracy of interpretation of DGA results.

The result of the prediction done to test the accuracy of the method shows the improvement of accuracy with 9% compared with the simple Key Gas Method.